Skip to main content
Vector Logo

​
Welcome to Vector

Vector is a high-performance, end-to-end observability data pipeline that puts you in control of your observability data. Built in Rust, Vector enables you to collect, transform, and route all your logs, metrics, and traces to any vendors you want today and any other vendors you may want tomorrow.

Quickstart

Get Vector up and running in minutes

Installation

Install Vector on your platform

Configuration

Learn how to configure your pipeline

Deployment

Deploy Vector as an agent or aggregator

​
Why Vector?

Vector solves critical observability challenges through its unique combination of performance, reliability, and flexibility.

High Performance

Built in Rust for maximum throughput and minimal resource usage. Process millions of events per second.

Vendor Neutral

Works with 100+ sources and sinks. Switch vendors without disrupting your workflows.

Reliable

Memory-safe architecture with end-to-end delivery guarantees. Never lose your data.

Unified Platform

Handle logs, metrics, and traces in a single pipeline. Simplify your observability stack.

Powerful Transforms

Use VRL (Vector Remap Language) to parse, filter, and enrich your data on the fly.

Production Ready

Trusted by Atlassian, T-Mobile, Comcast, Discord, and many more organizations.

​
Key Features

Vector supports a vast ecosystem of integrations including:
  • Sources: file, kafka, syslog, http, docker_logs, journald, prometheus, and more
  • Transforms: remap, filter, aggregate, reduce, log_to_metric, dedupe, and more
  • Sinks: elasticsearch, s3, kafka, datadog, prometheus, clickhouse, and more
VRL is a purpose-built language for transforming observability data. It’s:
  • Fast: Compiled for maximum performance
  • Safe: Fail-safe by default with compile-time error detection
  • Ergonomic: Designed specifically for log and metric manipulation
Deploy Vector in multiple roles:
  • Agent: Lightweight data collector running on each host
  • Aggregator: Centralized data processing and routing layer
  • Hybrid: Combine both roles for complex topologies
Vector provides:
  • Memory-safe Rust implementation
  • End-to-end acknowledgments and delivery guarantees
  • Disk-based buffering for backpressure handling
  • Built-in health checks and monitoring

​
Quick Example

Here’s a simple Vector configuration that reads logs from a file, parses them, and sends them to Elasticsearch:
vector.yaml
sources:
  my_logs:
    type: file
    include:
      - /var/log/app/*.log

transforms:
  parse_logs:
    type: remap
    inputs:
      - my_logs
    source: |
      . = parse_json!(.message)
      .timestamp = to_timestamp!(.timestamp)

sinks:
  elasticsearch:
    type: elasticsearch
    inputs:
      - parse_logs
    endpoint: "https://elasticsearch.example.com"
    index: "logs-%Y.%m.%d"

​
Next Steps

Core Concepts

Understand Vector’s architecture and data model

Configuration Guide

Learn how to configure sources, transforms, and sinks

VRL Guide

Master the Vector Remap Language for data transformation

API Reference

Explore Vector’s GraphQL API for runtime management

​
Community & Support

Vector is maintained by Datadog’s Community Open Source Engineering team with contributions from a vibrant open source community.

GitHub

Star us on GitHub and contribute to the project

Community Chat

Join our Discord community for help and discussions

Build docs developers (and LLMs) love

Get started for freeTalk to us